Time for a definitively tangential post.
Recently I was ruminating on what it is to be intelligent, and how we judge something to be conscious. I was doing this because it seams that we may soon if not already have enough computing power to model a human brain. Even if we are not there yet Moors law still seams to be active. We also don't need the full power of a brain because the simulation can be run slower than realtime. i.e. 1 day of computer work gives us 1 hour of brain behaviour.
This may one day allow us to anwser questions about the nature of human intelligence
Is consciousness an emergence phenomenon?
My reasoning is if a computer can simulate a consciousness then it emerged from the 1’s and 0’s dancing to their simple rules. With computers 10, 1000, 1000000 times more powerful than now it might not be possible to reproduce consciousness. In that outcome we might have to conclude that intelligence is fundamentally different from raw processing power. Not just because we have emotions, our super super computer would be able to model the chemical soup with think in perfect well. Note there is no cut off point where we can prove that consciousness is NOT emergent but with each failed attempt we gain evidence to that point. I think it is very likely in my life time we will have strong confidence on the answer and wonder what this might mean for us. Admittedly we will probably just carry on because hardly anyone read the paper and the media will missreport it.
Next up.
How can we tell if a computer is intelligent?
The Turing test says that a computer is intelligent when it can fool a person into believing it is human. Ready for a segway? what is sleep for, it must be fundamental to managing a biological intelligence. Why? because if not it would have been evolved out. Sleeping is not a good survival tactic. So my test for intelligence is as follows.
A machine is intelligent when, without prompting, it asks for a nap.